Array is incorrectly shifted for 96x16 OLEDs. about luma.oled HOT 5 CLOSED

So it will be already correct for the vast majority of devices .. except yours. I have found there to be many quirks with what are the "same" chipsets.. so the key here would be to retain existing behaviour but allow that parameter to be overridden for cases such as yours.

from luma.oled.

Are there 96x16 devices in the wild that follow the existing _colstart math? If so, then definitely, those devices absolutely need to keep working, and I can draft a PR that allows overrides.

from luma.oled.

Actually, I think a regression with 96x16 displays was introduced in 2017, in this commit and my PR should correct it.

The 2017 commit added the _colstart logic in question. Before it, 96x16 displays were supported but would have had an equivalent _colstart of 0. After it, 96x16 displays would have used a _colstart of 16. My PR restores the original behaviour.

from luma.oled.

the key here would be to retain existing behaviour but allow that parameter to be overridden for cases such as yours.

Agreed. Ideally, this would be tested with (eg intel desktop machine) and without ftdi_i2c (rpi).

from luma.oled.

@thijstriemstra - I can put together something that exposes _colstart as a parameter if you like, but I think this issue (and the linked PR #368) are pure regression fixes that do not justify a configurable parameter or API changes on their own.

ed: in case this seems fussy: I want to be sure I'm not adding clutter code to a project I may not use long-term.

from luma.oled.